Late one afternoon in the middle of March, a gray whale dives in Penn Cove off Whidbey Island. On the right underside of its tail fluke, three large, irregularly shaped pale spots mark this whale as CRC2356 or Stalwart, one of a dozen or so gray whales known as "Sounders" who descend on Puget Sound each spring to feast on ghost shrimp. The whales scoop great mouthfuls of mud from shallow tidal flats and force it out through their baleen plates, gulping down the tasty four-inch crustaceans that remain.

In front of the whale and a little to the west lies the grid of wooden rafts belonging to Penn Cove Shellfish, which at this time of year will be hung with lines coiled at the top of the water to catch mussel spat from the spring spawn. Penn Cove is one of the most productive shellfish-growing locations on the U.S. West Coast, rich in nutrients that fuel the growth of the plankton that mussels and other filter-feeders depend on. 

Beneath this idyllic scene, at the bottom of the cove on the seafloor, water trickles into the cove from the Pacific Ocean. The ever-so-faint current might be barely registered by the knobby papillae on the bodies of sea cucumbers that hang out below the rafts and feed on the bonanza of mussel waste that sifts down from above.

The ocean water has traveled along the Strait of Juan de Fuca, through Admiralty Inlet and around the southern tip of Whidbey Island then up through Saratoga Passage, sluicing through glacier-carved channels and tipping headlong over sills to reach Penn Cove. Its arrival will establish how well the local ecosystem “breathes” and whether life can continue to thrive here. This springtime inflow will shape dissolved oxygen (DO) levels in the cove until sometime in the fall.

It is a simple fact that almost all living things need oxygen to survive — from whales to shellfish — and diminished levels of dissolved oxygen can make marine organisms more vulnerable to disease, interfere with their sensory perception, impair their reproduction, and may even be fatal. In Penn Cove, for example, oxygen levels at the bottom of the water can fall as low as 0.18 mg/L in the late summer and early fall.

Among other consequences, low oxygen may weaken mussels’ byssal threads, which they use to attach themselves to the rocks, ropes, or other substrates on which they live. Although there have been no dramatic consequences of low oxygen in Penn Cove so far, such as fish kills that have occurred in some parts of Puget Sound, scientists are closely monitoring trends in oxygen levels here and elsewhere as the climate warms and pressures from human development continue. 

This is why scientists consider dissolved oxygen to be a critical measure of an area’s water quality, and it’s why a pair of sleek, white, two-foot-long cylinders are attached to a piling on the Coupeville Wharf. The cylinders contain sensors that log water temperature, salinity, pressure, pH, dissolved oxygen, chlorophyll, turbidity, and nitrogen every 15 minutes as part of a King Country program to monitor water quality in the Whidbey Basin of Puget Sound.

The scene at Penn Cove is the backdrop to an effort to understand which parts of Puget Sound suffer from low levels of dissolved oxygen, when, and why. Problems with dissolved oxygen in Puget Sound vary both by location and by season, this research is revealing. But many questions remain – and the answers could have multi-billion-dollar stakes.

Elements of the problem

Penn Cove, like many parts of Puget Sound is naturally prone to low oxygen due to circulation patterns and bathymetry, or the shape of the underwater terrain. The question of what exactly constitutes natural conditions in Puget Sound has been contentious. But the broad strokes of oxygen dynamics are well understood. Waters of different temperature or salinity tend not to mix, so fresher, more oxygenated water pouring into Puget Sound from rivers tends to sit on top of saltier, oxygen-poor marine waters from the Pacific Ocean. This phenomenon, known as stratification, worsens the depletion of oxygen in bottom waters. Long, narrow, and shallow bays and inlets are especially prone to this issue because their waters are slow to be exchanged with water from larger water bodies. Penn Cove, which pushes more than three miles into the eastern shore of Whidbey Island like a finger making a deep dimple in soft bread dough, is a typical example.

While some parts of Puget Sound are prone to low oxygen year-round, the problem tends to be worse and more widespread in late summer and fall. This is partly because warm weather during summer makes stratification more pronounced. In addition, summer’s long, sunny days drive the growth of photosynthetic organisms including microscopic cyanobacteria and algae. When the blooms die back, the remains of these organisms fall to the bottom and are broken down by bacteria, a process that consumes oxygen.

Human activities can also worsen oxygen depletion, chiefly by increasing the level of nutrients, especially nitrogen, entering Puget Sound. Excess nitrogen comes from agricultural runoff, fossil fuel burning, and especially discharges from wastewater treatment plants serving the region’s burgeoning human population. 

The large majority of human-caused nutrient pollution comes from wastewater treatment plants, according to Colleen Keltz, Communications Manager for the Water Quality Program at the Washington State Department of Ecology. "[It] doesn't mean they generate the pollution," she says. “It means we, as humans, pee and flush the toilet.” When that happens, nitrogen-rich effluent from wastewater treatment plants can act as a fertilizer, turbocharging summer algal blooms and making the subsequent fall in dissolved oxygen more acute.

Most of the nitrogen found in Puget Sound comes from natural sources, the vast majority of that from the ocean. An estimated 9% of the nitrogen can be traced to human activities in Washington State. While the anthropogenic contribution is small overall, nitrogen is normally a limiting nutrient in Puget Sound waters, so just a little extra may result in substantial overgrowth of phytoplankton blooms and subsequent lack of oxygen. 

Just how much that amount might tip the balance has become a matter of debate that is driving scientific research and has set off a series of legal battles over what should be done about it.  

Areas of concern

Excess human-caused nutrients are a widespread problem affecting waterways worldwide. Nutrient reduction strategies have yielded success in cleaning up the Chesapeake Bay and waters around Denmark. In Puget Sound, some Native tribes and local communities, especially in areas with waters affected by low oxygen, argue that technology to remove nutrients from wastewater should be added to treatment plants around the region. But this strategy could cost billions of dollars, costs that would be passed on to utility ratepayers. And it’s important to know exactly where nutrient pollution needs to be reduced and by how much.

Some guidance on that question comes from water quality standards. The Washington State Department of Ecology regulates dissolved oxygen levels in marine waters as part of its responsibility for implementing the federal Clean Water Act. Standards range from 4 to 7 milligrams of oxygen per liter, depending on the needs of the species that live (or could live) in a particular location. For most of Puget Sound this so-called biologically based or numeric standard is 6 or 7 mg/L; in Penn Cove it is 6. 

Waters can fall afoul of state dissolved oxygen standards without being hypoxic, a term that most commonly indicates dissolved oxygen levels of less than 2 mg/L. But state officials say these higher standards are necessary to protect the health of aquatic life. “These are the oxygen concentrations that we know the fish that are existing in these specific parts of Puget Sound and other critters need to thrive,” says Jeremy Reiman, an environmental planner at the Department of Ecology working on water quality standards.

Washington State is currently in the process of re-establishing water quality standards based on the dissolved oxygen levels that would have historically, prior to European settlement, prevailed at a particular location. Once such “natural conditions” criteria are in place, they will replace the numeric standard wherever they apply. Under the natural conditions regime, state regulations include an “anthropogenic allowance” for human impacts. Human activities cannot reduce dissolved oxygen by more than 0.2 mg/L or 10% (whichever decrease is smaller).

A primary tool for understanding the human influence on oxygen levels in Puget Sound is the Salish Sea Model (SSM), a state-of-the-art computer simulation developed by the Pacific Northwest National Laboratory in collaboration with the Washington Department of Ecology. The SSM can also provide a more detailed, high-resolution picture of current oxygen conditions than is available from on-the-water monitoring data – and thus point to areas where greater monitoring efforts are needed.

In one SSM analysis, the model revealed roughly a dozen and a half areas around Puget Sound that may be out of compliance with state dissolved oxygen standards due to human activities and where more information is needed about local oxygen dynamics. The areas of concern include parts of Hood Canal, a long, narrow, deep arm of the sea that has what researchers in a 2018 paper on hypoxia in Puget Sound dubbed a “classic fjord-type circulation” with strong stratification and a “nearly stagnant” bottom layer, and regularly becomes depleted of oxygen in the summer. The list also includes many shallow terminal bays and inlets around the region. Some of the places on the list have long been known as areas prone to low oxygen. Budd Inlet, in the far southern reaches of Puget Sound near Olympia, has been the subject of nutrient cleanup efforts since the 1990s. But the list also includes some new areas of concern, including Penn Cove. The analysis provides a focus for concern about low oxygen levels in Puget Sound: the problem isn’t everywhere, says Stefano Mazzilli, senior research scientist at the University of Washington Puget Sound Institute who has been working with the model. “There is only a finite number of places,” he says, “So we can start by focusing on what’s driving change in those places.” [Editor's note: The Puget Sound Institute is the parent organization of the Encyclopedia of Puget Sound.]

A tale of two models

More clues about when, where, and why low oxygen levels occur in Puget Sound come from a second computer model, known as LiveOcean. An analysis presented at the Science of Puget Sound Water Quality workshop series in February focused on six inlets in Puget Sound where LiveOcean predicts that average daily oxygen levels at the bottom of the water fall below 2 mg/L – that is, become hypoxic – during the late summer and early fall. That list includes Penn Cove. 

Researchers analyzed these six hypoxic inlets compared to seven better-oxygenated inlets in Puget Sound to understand what processes contribute most to the development of low oxygen levels during August and September. According to the prevailing model of coastal oxygen dynamics, hypoxia might develop in a particular area because the water there starts out with relatively low oxygen levels, that is, at the beginning of the growing season in spring; because biological processes such as the breakdown of phytoplankton blooms deplete oxygen quickly; or because water sticks around in the area for a long time before being replaced with more oxygenated water from elsewhere.

The first and third of those factors are the main determinants of susceptibility to hypoxia in Puget Sound’s inlets, the analysis showed. “The most hypoxic inlets in Puget Sound are not ones that act as isolated hotspots of high DO consumption or high DO depletion rates,” Aurora Leeson, a graduate student in the laboratory of University of Washington oceanographer Parker MacCready, which developed and maintains LiveOcean, told the workshop. “Instead, deep DO concentrations in terminal inlets of Puget Sound are strongly influenced by what flows into the inlets and how long it takes to flush the inlet.” 

This analysis is based on 2017 data and represents current conditions, Leeson says. A next step is to analyze a reference scenario without anthropogenic nutrient inputs, then compare the two to determine how wastewater treatment plants and other human activities may be contributing to the dynamics that produce hypoxia in these inlets.

The two models have many differences: SSM divides Puget Sound into a landscape of triangles, while LiveOcean uses a grid of squares; SSM offers greater resolution at the surface of the water while LiveOcean divides the depth of the water column into finer layers; SSM focuses on minimum oxygen concentrations to LiveOcean’s tidally-averaged daily values; and so on.

But while some details of the analyses conducted and results obtained may differ, in broad strokes the two models paint a similar picture of the problem of hypoxia in Puget Sound: The areas most affected are Hood Canal and many shallow bays and inlets with long flushing times. But the causes of hypoxia can’t be localized just to those affected areas; “Puget Sound is so interconnected,” Leeson told Salish Sea Currents.

This means it’s necessary to reduce nutrient inputs not just near hypoxia hotspots but throughout Puget Sound, says Ecology’s Colleen Keltz. The agency has sought to incorporate these insights into regulations with its Puget Sound Nutrient General Permit, issued in December 2021 and covering 58 wastewater treatment plants that discharge into different parts of Puget Sound. 

But on February 28, the Washington State Pollution Control Hearings Board struck down this approach, ruling that the general permit cannot be a mandatory overlay on existing individual permits. Ecology decided not to appeal the ruling and is shifting to a voluntary general permit approach.

Trends and tinkering 

Meanwhile, efforts continue to both refine the models and build a better understanding of actual conditions in the water. Shallow terminal embayments – precisely the sorts of areas that are prone to low oxygen levels – are also difficult to render precisely in large-scale ocean models, says Taylor Martin, an oceanographer with King County.

Spurred by SSM results suggesting that nutrient discharges from King County wastewater treatment plants could contribute to low oxygen levels in Penn Cove and nearby areas, the county has launched a water quality monitoring program in the Whidbey Basin. “A lot of the water that comes from King County flows back North,” Martin explains. “A lot of it goes back out Admiralty Inlet. But not all of it. Some of it does slosh around to the East side of Whidbey.” 

King County began sampling water quality at 10 locations in Whidbey Basin in 2022, initially collecting data twice and now once monthly. In 2023 they installed continuous monitoring instruments at three locations – near Port Susan, near the mouth of Penn Cove, and at Coupeville Wharf – that record data every 15 minutes. 

So far in Penn Cove, “I think we've seen basically exactly what we expected to see,” Martin reports. “Dissolved oxygen does get quite low in the late summer and fall.” The county plans to continue monitoring these areas to see if there are any trends in the timing or duration of hypoxia – which can be difficult to establish because patterns of dissolved oxygen can vary so much from year to year, Martin says.

The Department of Ecology has been continuing to tinker with the SSM to find the best strategies for nutrient reduction, looking for scenarios in which the model predicts the highest compliance with dissolved oxygen standards throughout Puget Sound. Its modeling team teased the latest results at a meeting of the Puget Sound Nutrient Forum on March 27, and the agency plans to release the results along with an Advance Restoration Plan to reduce nutrients in Puget Sound in June.

This article was funded in part by King County in conjunction with a series of online workshops exploring Puget Sound water quality. Its content does not necessarily represent the views of King County or its employees.